1 Setup

library(tidyverse)    # to do datascience
library(INBOtheme)    # to apply INBO style to graphs
library(sf)           # to work with geospatial vector data
library(plotly)       # to make dynamic plots
library(leaflet)      # to make dynamic maps

2 Introduction

In this document we will:

  1. read occurrence cube data
  2. explore data
  3. visualize data

3 Read data

Read ABV data from the occurrence cube file 20251028_abv_cube.csv:

abv_cube <- readr::read_csv(
  file = "./data/20251028/20251028_abv_cube.csv"
)
## Rows: 1251 Columns: 17
## ── Column specification ────────────────────────────────────────────────────────
## Delimiter: ","
## chr (8): mgrscode, species, order, family, genus, id_dataset, id_spat_res, c...
## dbl (9): year, specieskey, n, mincoordinateuncertaintyinmeters, ordercount, ...
## 
## ℹ Use `spec()` to retrieve the full column specification for this data.
## ℹ Specify the column types or set `show_col_types = FALSE` to quiet this message.

Read the Flemish grid from the geopackage file 20251028_utm_grid.gpkg:

vl_grid <- sf::st_read("./data/20251028/20251028_utm_grid.gpkg")
## Reading layer `utm_grid' from data source 
##   `C:\R\git_repositories\coding-club\data\20251028\20251028_utm_grid.gpkg' 
##   using driver `GPKG'
## Simple feature collection with 14221 features and 4 fields
## Geometry type: POLYGON
## Dimension:     XY
## Bounding box:  xmin: 21104.48 ymin: 152496.5 xmax: 259486.4 ymax: 244558.5
## Projected CRS: BD72 / Belgian Lambert 72

4 Explore data

This dataset contains data from 2007 to 2023 related to 3 species from the Emberizidae family and their distribution in Flanders based on a grid of 1 km x 1 km.

Preview of the first 30 rows of the dataset:

head(abv_cube, n = 30)

4.1 Taxonomic information

Species present in the dataset:

abv_cube %>% distinct(specieskey, species)

4.2 Temporal information

The data are temporally defined at year level. Years present:

abv_cube %>% dplyr::distinct(year) %>% arrange(year)

4.3 Geographical information

The geographical information is represented by the mgrscode column, which contains the identifiers of the grid cells containing at least one occurrence of the species.

The dataset contains 382 unique grid cells.

5 Preprocess data

Add geometrical information to the occurrence cube via mgrscode, which contains the identifiers of the grid cells containing at least one occurrence of the species.

cells_in_cube <- vl_grid %>%
  dplyr::filter(mgrscode %in% unique(abv_cube$mgrscode)) %>%
  dplyr::select(-c(TAG, Shape_Leng, Shape_Area))
sf_abv_cube <- cells_in_cube %>%
  dplyr::left_join(abv_cube, by = "mgrscode")

Final (spatial) dataset:

sf_abv_cube %>% head(n = 30)

6 Data visualization

In this section we will show how the number of occurrences and the number of occupied grid cells vary by year and species. Both static plots and dynamic maps are generated.

6.1 Static plots

6.1.1 per species

n_per_species <- sf_abv_cube %>%
  dplyr::group_by(species) %>%
  dplyr::summarize(occurrences = sum(n),
                   grid_cells = n_distinct(mgrscode),
                   .groups = "drop") %>%
  tidyr::pivot_longer(cols = c(occurrences, grid_cells),
                      names_to = "variable",
                      values_to = "n")

ggplot(n_per_species, aes(x = species, y = n)) +
  geom_bar(stat = 'identity') +
  facet_grid(.~variable, scales = "free_y") +
  ggplot2::theme(axis.text.x = element_text(angle = 60, hjust = 1))

6.1.2 per year

n_per_year <- sf_abv_cube %>%
  dplyr::group_by(year) %>%
  dplyr::summarize(occurrences = sum(n),
                   grid_cells = n_distinct(mgrscode),
                   .groups = "drop") %>%
  tidyr::pivot_longer(cols = c(occurrences, grid_cells),
                      names_to = "variable",
                      values_to = "n")

ggplot(n_per_year,aes(x = year, y = n)) +
  geom_bar(stat = 'identity') +
  facet_grid(.~variable, scales = "free_y") +
  ggplot2::theme(axis.text.x = element_text(angle = 60, hjust = 1))

6.1.3 per year and species

n_occs_per_year_species <-
  sf_abv_cube %>%
  dplyr::group_by(year, species) %>%
  dplyr::summarize(occurrences = sum(n),
                   grid_cells = n_distinct(mgrscode),
                   .groups = "drop") %>%
  tidyr::pivot_longer(cols = c(occurrences, grid_cells),
                      names_to = "variable",
                      values_to = "n")

ggplot(n_occs_per_year_species,
       aes(x = year, y = n, fill = species)) +
  geom_bar(stat = 'identity') +
  facet_grid(.~variable) +
  ggplot2::theme(axis.text.x = element_text(angle = 60, hjust = 1))

6.2 Dynamic plots

6.2.1 Leaflet dynamic map

We show a map with the distribution of buntings in Flanders (??). We show the total number of occurrences per grid cell. The color of the grid cells is based on the number of occurrences. The legend shows the color scale and the number of occurrences per grid cell.

n_occs_per_cell <- sf_abv_cube %>%
  dplyr::group_by(mgrscode) %>%
  dplyr::summarize(
    occurrences = sum(n),
    min_coordinateuncertaintyinmeters = min(mincoordinateuncertaintyinmeters),
    .groups = "drop")

map_abv <- mapview::mapview(n_occs_per_cell,
                            zcol = "occurrences",
                            legend = TRUE
)

map_abv

6.2.2 Plotly yearly abundance

We show a graph with the yearly abundances per species (add reference to figure here).

n_occs_per_year <- n_occs_per_year_species |>
  dplyr::filter(variable == "occurrences") |>
  st_drop_geometry()

fig <- plot_ly(n_occs_per_year,
               x = ~year,
               y = ~n,
               split = ~species,
               stroke = ~species,
               type = "scatter",
               mode = "lines+markers")

fig